Previous day | Jump to hour: 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 | Next day

Seconds: Show Hide | Joins: Show Hide | View raw
Font: Serif Sans-Serif Monospace | Size: Small Medium Large

Click in the nick column to highlight everything a person has said.
The Logo icon identifies that the person is a core developer (has commit access).

#rockbox log for 2004-11-22

00:03:39s0cksHi midk
00:03:42s0cksHow goes?
00:05:56*Bagder heard Linus now has the FAT code working on the iriver
00:06:13s0cksFAT is good.
00:06:23s0cksWhy is he trying to partition a flash player?
00:06:44s0cksFile Allocation Table?
00:07:01BagderFAT32 == file system
00:07:50***Saving seen data "./dancer.seen"
00:08:15s0cksfat = file system also.
00:08:59Bagderwell, we use the flavour of fat often called fat32
00:09:31Bagderexcept on the ondio where we use fat16 (too?)
00:10:09amiconnBagder: yes.
00:10:33amiconn(Actually MMCs are usually FAT16, but FAT32 is also supported)
00:11:32s0cksSweet. Are they coming out with fat64? or has everyone left fat for something else?
00:15:39Bagderwe don't have to care about that
00:16:52s0cksJust wondering
00:27:45 Join kurzhaarrocker [0] (
00:38:34 Quit _aLF ("Leaving")
00:38:35 Quit kurzhaarrocker (Read error: 104 (Connection reset by peer))
00:48:29 Quit midk ("just STOP it arspy")
00:50:05 Join midk [0] (
00:50:41 Nick AciD is now known as AciD` (
00:51:09 Nick AciD` is now known as AciD (
00:52:40 Nick AciD is now known as AciD` (
00:52:49 Part s0cks ("Leaving")
00:52:52 Nick AciD` is now known as AciD (
00:54:10 Quit adi|ems (Read error: 110 (Connection timed out))
00:59:24 Quit AciD (Read error: 104 (Connection reset by peer))
01:46:37 Quit midk ("just STOP it arspy")
01:51:42 Join gromit`grosse`MA [0] (
01:51:42 Quit gromit (Read error: 54 (Connection reset by peer))
01:51:43 Quit gromit`grosse`MA (Remote closed the connection)
01:52:03 Join gromit [0] (
01:52:13 Part gromit
02:07:51***Saving seen data "./dancer.seen"
02:38:30 Join amiconn_ [0] (
02:43:03 Quit amiconn (Nick collision from services.)
02:43:03 Nick amiconn_ is now known as amiconn (
02:52:39 Join adi|ems [0] (
02:53:00 Quit adi|ems (Client Quit)
03:03:51 Part amiconn
03:50:32 Join midk [0] (
04:07:52***Saving seen data "./dancer.seen"
05:55:54 Quit Stryke` ("Friends don't let friends listen to Anti-Flag")
06:07:55***Saving seen data "./dancer.seen"
06:46:16 Quit midk ("just STOP it arspy")
07:13:29 Join AciD [0] (~gni@acid.user)
07:28:28 Join gromit`_ [0] (
07:31:31 Join midk [0] (
07:31:33 Join LinusN [0] (
07:38:18dwihnoWelcome \o/
07:38:26dwihnoI hope you did enjoy your weekend.
07:41:48LinusNoh yes
07:45:02 Quit gromit` (Read error: 110 (Connection timed out))
07:48:45 Join oxygen77 [0] (
08:00:52 Part oxygen77 ("Cho")
08:07:56***Saving seen data "./dancer.seen"
08:26:25 Join Bagder_ [0] (
08:30:01LinusNmorning Bagder
08:30:04LinusNmorning Bagder_
08:30:09Bagder_good morning
08:30:27LinusNwhat's the deal with your nicks?
08:30:29Bagder_the non underscore me is my upstairs computer
08:30:50 Join amiconn [0] (
08:31:01amiconngod morgon
08:31:21LinusNBagder_: i found out the ATA power/reset connections
08:31:43LinusNamiconn: guten morgen (?)
08:32:03amiconncorrect :)
08:32:12LinusNBagder_: we were sooo wrong :-)
08:32:32LinusNamiconn: i learnt german from the 64'er magazine
08:32:48LinusNbut guten morgen was a seldom used phrase there :-)
08:32:54amiconnLinusN: Maybe there was a reset-finding day? (If you noticed my finding on the player...)
08:33:02LinusNamiconn: :-)
08:33:38LinusNgood work on the player, btw. my hat is off
08:34:12LinusNtime to replace my flash chip
08:34:23amiconnA propos player: It would be very kind if you could do some tests on your old player
08:34:26LinusNBagder_ i could do yours as well
08:34:41Bagder_it'd be great
08:34:47LinusNamiconn: will do. any special things you'd want me to test?
08:35:08amiconnMainly 2 things (if you don't change the flash chip yet):
08:35:40amiconn(1) Try to uart-boot a current archos image (5.08) to see if this works with the old hardware
08:36:20amiconn(2) Try to uart-boot current rockbox mainy to check the lcd init
08:36:33amiconnFurthermore I would be interested in a ROM dump
08:36:59amiconnBtw: Which ROM version does it have?
08:37:36LinusNhmmm, 3.xx something, don't remember
08:40:05amiconnIf you want to solder in a flash chip, this would be probably the best test, but it may be that you would still need the uart boot mod, in case neither the archos image nor rockbox do work
08:43:05amiconnI currently have a test running: a rockbox build with ata power off on the player. It's already running for ~6 hours now, continually playing music. No crash yet :)
08:43:48amiconnDo you roughly remember when you did those tests (as you reported occasional crashes)?
08:46:00LinusNback when jesus had diapers, i guess :-)
08:46:27amiconnBtw: This is even with a Hitachi DK23DA-10 disk
08:46:33Bagder_I knew you are old LinusN, but not _that_ old ;*)
08:48:03amiconnLinusN: You fixed the init after reapplying ata power for the recorder back in February. Maybe the same problem was also the cause for the crashes on the player, if your tests were done before that
08:48:44LinusNit was long before that, so that may be it
08:49:02LinusNthose test were made back in 2002 methinks
08:49:31*LinusN will try it on his newplayer immediately
08:50:17amiconnIf you want to test, you merely have to set a #define in config-player.h now. Everything else is already prepared :)
08:50:20LinusNbtw, the old player doesn't have ATA power control
08:50:47amiconnI know, enabling this for everyone would need to dynamically show/ hide the menu entry
08:51:58LinusNand not touch PB4 on old players
08:52:04amiconn..and set the global settings entry to "no" for old players, although setting it to "yes" does no harm, the power functions check for new/ old
08:52:33amiconnSee my power.c/ power.h commits :)
08:54:58LinusNah, an alias...
08:55:20amiconnYes, saving some code...
08:55:29LinusNinstead of a macro...?
08:56:43amiconnA macro would save the same amount of code... only it gives the original function name instead of the aliased one compiler in warnings/ error
08:57:15amiconnWe could also rename this function to is_new_player() or such, and get rid of aliasing
08:59:13LinusNmuch better
09:03:05LinusNtested disk poweroff on my newplayer
09:03:11LinusNsame problems as last time :-(
09:03:35 Quit Bagder_ ("Leaving")
09:03:40LinusNplayer shuts off when trying to power on the hard drive
09:05:03LinusNmaybe this is a problem specific to my player?
09:05:28LinusNsomeone else should try it
09:05:48amiconnHmm, maybe. It can't do that always, otherwise it would switch off right after poweron...
09:07:43amiconnWhat ROM version/ PCB version(s) does your new player have? Maybe this is hw version dependent
09:09:15LinusNfw 4.52
09:09:20LinusN4.53 sorry
09:09:24amiconnDid you reset the settings after loading your test build? The settings structure changes, and some settings may have gotten weird values
09:10:35LinusNi did
09:11:31amiconn(away now)
10:07:57***Saving seen data "./dancer.seen"
10:14:26 Join Zagor [242] (
10:14:47Bagdermorning Z
10:27:40 Quit AciD (Read error: 104 (Connection reset by peer))
10:28:15amiconnLinusN: The fw version on your newplayer - is that the ROM version, or the version of the archos image (they may be different)?
10:28:16 Join AciD [0] (
10:29:02LinusNamiconn: of course it is the ROM version, i have no archos image as i run rockbox
10:29:55LinusNor have i missed something?
10:30:36amiconnOf course your ROM contains an archos image.
10:31:15amiconnThe ROM version is what rockbox displays under debug->view HW info, versus the archos image version is what it displays at boot
10:31:27amiconnMy ROM is 5.06, versus the archos image is 5.08
10:31:41LinusNnow i get it
10:31:44LinusNlemme see
10:34:39amiconnWhich one?
10:36:56amiconnHmm. I still wonder what may cause those weird shutdowns with ata power off/ on. At least your newplayer is one of the early newplayers. Maybe there are some hw problems that were solved later
10:37:29 Join MooMaunder [0] (~me@
10:39:08amiconnLinusN: Do you have access to any other newplayers, were you could test on?
10:40:44LinusNmatsl has one, as well as kjer
10:47:22 Join kurzhaarrocker [0] (
11:04:06 Join bobTHC [0] (
11:04:10bobTHChi all!
11:10:37bobTHCalways under the snow ??
11:11:16LinusNtoday's forecast says that it will melt away today... :-(
11:11:35Zagorfinally we'll get traditional swedish winter ;)
11:12:56bobTHCthe winter god retakes his right, good news for north ppl ;)
11:13:25*kurzhaarrocker wants a plugin to ffwd / rewind seasons
11:14:31LinusNi *want* snow and cold
11:15:33kurzhaarrockerI don't mind you having snow and cold as long as you're far away with that.
11:20:41 Join Digital007 [0] (
11:21:19Digital007i don't think iriver are gonna make new firmwares for the H1xx series
11:21:26Digital007they want to concentrate on the H3xx series
11:23:29LinusNand you realize that *now*? :-)
11:23:36Digital007from misticriver
11:23:47Digital007I hope rockbox is introuduced on iriver soon
11:24:01Digital007i really do
11:24:35Digital007despite its shuffle foibles i still think its a far better machine than the Archos
11:26:21kurzhaarrockerbetter than the archos recorders? That's not difficult :)
11:26:41Digital007ive got an H120
11:27:00Digital007well i suppose the Archos has those snazzy blue corners which the iriver lacks :-)
11:27:29amiconnLinusN: Do you have the original hi-res scans of the archos models somewhere?
11:28:27Digital007Linus when is the approximate release period for iriver rockbox?
11:28:48LinusNamiconn: we don't have any other scans that those on the site
11:29:15kurzhaarrockerDigital007: I prefer the black corners - I think they make the sound a bit more smooth
11:30:01LinusNamiconn: ah, you mean the ones i did for the model overview?
11:30:06amiconnLinusN: I presume the scans on the site to be scaled down
11:30:33Digital007Are colour screens on an MP3 player really necessary?
11:30:48bobTHCimho not
11:30:57Digital007i hate colour screens on MP3 players
11:31:04LinusNDigital007: i thought we have made it very clear that we don't have a release schedule
11:31:15Digital007well i don't think u did linusn
11:32:24Digital007like that photo ipod (not)
11:34:48LinusNDigital007: 23.53.01 # <Digital007> when u think the first release will be?
11:34:48LinusN23.53.23 # <LinusN> when it's ready
11:34:48LinusN23.53.32 # <LinusN> not this year
11:35:14Digital007sorry i forgot
11:35:56LinusNi agree that my reply wasn't clear about not having a schedule at all
11:36:29LinusNbut since you frequent misticriver i thought you would have followed the "Rockbox progress" thread
11:36:39Digital007i didn't know there was one
11:37:16Digital007too busy hammering iriver for a fix to the shuffle prob
11:37:23LinusNi always select "New Posts" in the menu
11:37:57LinusNdamn Freescale for their sucky data sheets
11:46:16chendowow, this channel is finally active
11:46:36chendowell, active enough by my standards
11:46:50LinusNit's been active since march 2002 :-)
11:46:50chendohow do you guys reflash the firmware if you happen to screw it up?
11:46:58chendowell, not since i was in here :P
11:47:03LinusNhow did you screw it up?
11:47:12chendowell, i didn't (yet)
11:47:16chendojust wondering how you guys do it
11:47:27chendosince you can only flash from a working firmware on the iRiver
11:47:45LinusNah, you're talking about iriver...
11:48:00chendoyeah... interested in seeing what you guys will make :)
11:48:20LinusNnone of us has flashed anything yet
11:48:20chendoalthough, are you just gonna leave it monochrome like on the archos or are you gonna prettify the GUI?
11:48:42LinusNwe will probably "prettify" it eventually
11:49:38chendobetter video playback? ;o
11:49:48LinusNthan the archos?
11:50:18chendowell, than what iRiver has now
11:50:42LinusNi didn't know that iriver had video playback on the H1xx
11:51:28chendowell, on the H3xx...
11:51:57bobTHCmaking rockbox "as it" working on the iriver devices is already a big challenge
11:52:19chendoyeah well, i consider myself an optimist :)
11:52:26 Quit Digital007 ("CGI:IRC (Ping timeout)")
11:53:13chendothere's a CGI:IRC? o.o
11:53:15LinusNas you probably know, we have few real plans for the iriver port, except porting it
11:53:24chendoaw :(
11:54:17LinusNonce it's ported, we'll see what happens next
11:55:45kurzhaarrocker"real plans"? Seems I missed something.
11:56:37LinusN"few real plans" really means "no real plans"
11:57:35LinusNlunch time
11:59:33bobTHCbut "no real plans" not really means "no ideas "
12:00:50kurzhaarrocker<- must learn reading. I read "we have _a_ few plans" instead of "we have few plans"
12:07:59***Saving seen data "./dancer.seen"
12:30:29 Join midk_ [0] (
12:30:30 Part LinusN
12:30:34 Part Zagor
12:31:19 Join LinusN [0] (
12:31:19 Join Zagor [242] (
12:33:36 Quit MooMaunder (Read error: 54 (Connection reset by peer))
12:46:43 Quit midk (Read error: 110 (Connection timed out))
13:17:08 Join [IDC]Dragon [0] (
13:20:41amiconnMorning? ;)
13:21:38[IDC]Dragonmeetings all morning
13:25:40 Join starmanager [0] (
13:28:32 Quit starmanager (Client Quit)
13:55:54Zagor"DMC Xclef 500 137GB Portable - Made in the USA" −−
13:56:49Zagoramazing how some companies choose to lie about their products
13:58:22Zagorthey "plan to release one". right. as soon as someone makes a disk that size...
13:58:38[IDC]DragonI should wait with my HD upgrade
14:00:39kurzhaarrocker137 GB? No way with my USB1.1 recorder.
14:08:01***Saving seen data "./dancer.seen"
14:59:15 Join casualtie [0] (
15:26:22 Join Headie [0] (
15:53:21 Part LinusN
15:56:39 Join gromit` [0] (
16:08:05***Saving seen data "./dancer.seen"
16:11:02 Quit casualtie (Read error: 110 (Connection timed out))
16:16:39 Join Tang [0] (
16:17:42 Join pfavr [0] (
16:22:26pfavrhi! I just replaced the drive in my recorder 20 with a 40GB hitachi, and installed rockbox 2.3. I did the partitioning using linux fdisk and chose partition type 'C' (fat32, lba). Then I formatted the drive using mkdosfs on linux. It seems to work fine - but "Rockbox Info" says the disk is only 3.10GB with 3.7GB free. Any ideas?
16:22:57Bagderrun that update disk space thing
16:23:20pfavra plugin?
16:23:21Zagorpress play in the disk space screen
16:23:29Zagor...and wait
16:24:06pfavrok, now I just need to find the disk space screen?
16:24:35pfavr(feeling a bit stupid)
16:24:42Zagordebug/disk info
16:25:11pfavrgot it
16:25:29pfavrwell, still 3846MB free
16:26:17pfavrBut it says size is 38154MB
16:26:22pfavr(which is fine)
16:26:39pfavrCluster size 4096 bytes
16:27:26pfavrIn the partition screen it shows one partition: 38154MB
16:27:48Bagderwell, 40GB formatted fat is likely to be just that
16:28:36pfavryes, it is fine - just wonder if the rockbox info is ok
16:28:50pfavr(or if I did something bad to the disk;-)
16:29:22Bagderso what is wrong do you say?
16:29:28Zagorif linux says the disk is fine, it is. sounds like rockbox gets confused for some reason.
16:29:51pfavrjust wondered if this was known/normal behaviour
16:30:04pfavrwhen using a 40GB disk
16:30:16Zagorusually it is fixed by recalculating as you did
16:30:18pfavror maybe I overlooked something
16:30:43Bagder40*1000*1000*1000/1024/1024 = 38146
16:31:19pfavruh oh, this is embarrasing - I seem to have made a too small filesystem
16:31:30pfavrshould have checked first before asking
16:31:52Zagordon't specify size, just let it use the defaults values
16:31:55pfavrdf -h says: /dev/sda1 4,0G 241M 3,8G 6% /mnt/archos
16:32:08Zagorsounds kinda small, yes :)
16:32:10pfavrI did use the defaults
16:32:30pfavrsounds like rockbox is reporting the accurate sizes
16:33:00pfavrthanks anyway
16:34:55 Quit bobTHC ("( :: NoNameScript 3.81 :: )")
16:35:31pfavrnow it works - it was probably because I didn't unplug and reconnect after doing fdisk - then it used the old partition size for mkdosfs
16:36:02pfavrdf -h now says: dev/sda1 38G 16K 38G 1% /mnt/archos
16:41:32pfavrRockbox info: Disk 37.2GB :-)
16:43:44 Join mecraw_ [0] (~lmarlow@
16:44:42 Part Zagor
17:25:06 Join atdj [0] (ju_@
17:26:58 Join casualtie [0] (
17:27:24 Quit pfavr ("ChatZilla 0.9.61 [Mozilla rv:1.7.3/20041007]")
18:03:21 Quit kurzhaarrocker (Read error: 104 (Connection reset by peer))
18:07:53 Join methangas [0] (
18:08:09***Saving seen data "./dancer.seen"
18:17:11amiconn[IDC]Dragon: r u there?
18:18:41TangHi folk!
18:19:04TangI have a question about lace LinusM progress
18:19:21Tanglast (not lace, sorry)
18:42:50[IDC]Dragonamiconn: now I am
18:46:10 Join _aLF [0] (
18:49:37TangBut he is not, lol$
18:49:54[IDC]Dragonseems so
18:50:37TangI would ask something about the iriverport
18:50:54Tangnot sure you could answer me...
18:51:36[IDC]Dragonneither do I, unless you ask...
18:51:51Tangok si i'll do
18:51:58Tangthe recent CVS report
18:52:09Tangsay that:
18:52:14Tang"iRiver: attempt to set up the SDRAM correctly"
18:52:32TangI was wondering if this SDRAMwere the "buffer" memory
18:52:50Tangor the firmware mmory
18:52:53[IDC]Dragonit's the main memory
18:53:25Tangmain memory means where is the firmware?
18:53:51[IDC]Dragonunless that runs directly from flash, yes
18:54:07Tangah okay i see
18:54:22Tangit's the SDRAM used by the processor
18:54:43[IDC]Dragonlike the RAM in your PC
18:54:47Tangbut not were is stocked the firmware when you upgrade
18:54:53Tangok i see
18:55:09Tangthanks a lot
18:55:48Tangin fact i was hesitating if it was the RAM opening access to alternative firmware
18:56:18TangAnyway any progress is good to hear ;)
18:56:34[IDC]Dragonno, just setting up the hardware to use the RAM (init, refresh, etc.)
18:57:47Tangokay it insat done just working on it
18:58:10Tangit isn't done yet just working on it
18:59:22[IDC]DragonI don't know, and it's most likely not your concern
19:01:14[IDC]Dragonthe iriver port is like a newborn baby moving it's legs for the first time, and people come in here and ask when it will b a dancer
19:01:43TangNo sorry it wasn't what i wanted to say
19:02:03Tangi meant
19:02:26TangLinusM is just working on the SDRAM
19:03:00Tangthe CVS notice dosent seem to indicate he suceeded on this point
19:03:10Tang(setting STDRAM)
19:03:32Tang(sorry i don't speak english very well)
19:04:25TangI wasn't asking any feature or something like that on the contrary
19:04:47[IDC]Dragonok ;-)
19:04:54[IDC]Dragonnever mind
19:05:14Tangok sorry for misundertsanding
19:07:15 Join matsl [0] (
19:09:30 Join Bruno [0] (
19:10:10Brunowow see even accesing ram is difficult on iriver!
19:10:37[IDC]Dragonthe Archos also needs a RAM init
19:10:54Brunoram init is ok
19:11:03[IDC]Dragonbut this is done by the boot code, not part of Rockbox
19:11:15Brunobut i've read linus code and you must use a workaround becasue of a bug!
19:13:13Brunoi must say :keep going linus!!
19:18:30 Quit Bruno ("Chatzilla 0.9.66c [Mozilla rv:1.7.5/20041108]")
19:23:47 Quit AciD (Read error: 104 (Connection reset by peer))
19:24:21 Join AciD [0] (~gni@acid.user)
19:25:07 Quit [IDC]Dragon ("CGI:IRC")
19:42:13 Quit gromit` (Read error: 104 (Connection reset by peer))
19:42:57 Join gromit` [0] (
19:44:31TangAnother thing but don't take in cosideration if it's kinda "confidential"
19:45:16TangI've seen on the board that a guy own aBDM and an iHP too
19:45:38Tanghe wasn't memeber of rockbox anyway...
19:46:07Tangis there any contact with him?
19:46:43Tang(fireblade20772 on your board)
19:47:50Tang( )
20:08:12***Saving seen data "./dancer.seen"
20:17:50 Join void [0] (
20:17:57 Nick void is now known as void_ (
20:18:03 Quit atdj (Client Quit)
20:18:09void_anyone around?
20:19:16void_I'm having a bit of trouble with my archos jukebox recorder 15, I've noticed that the battery indicator shows a question mark for a while after being fully charged and then after some use of it I get read access errors
20:19:50void_running rockbox-2.3
20:20:49 Quit casualtie (Read error: 110 (Connection timed out))
20:23:38amiconnvoid_: reg. the question mark: see
20:24:48amiconnread access errors might be caused by bad battery contacts (a common problem on the NiMH powered archos devices, though I didn't experience it myself yet)
20:27:56void_how can I tell?
20:31:15amiconnIt is reported that broken battery contacts sometimes also cause the recorder to crash/ reboot if the corners are squeezed. See for a short description/ repair guide
20:31:16 Join Stryke` [0] (~Chairman8@resnet-241-86.resnet.UMBC.EDU)
20:59:39 Quit AciD (Read error: 104 (Connection reset by peer))
21:06:30void_ok I have it open
21:06:39void_is there supposed to be a lead connector on both sides?
21:06:46void_I see one side has it
21:18:35 Quit Tang ("Chatzilla 0.9.66 [Mozilla rv:1.7.5/20041108]")
21:21:09 Join AciD [0] (~gni@acid.user)
21:25:52amiconnvoid_: From what it looks to me, both springs should be firmly soldered to their respective pads. The spring to the upper left (when the unit is in front of you in the normal position) has a wire running across it.
21:26:18amiconnThis wire is wrapped around the edge of the pcb, and should be soldered at both ends
21:27:15amiconn(sorry didn't notice earlier)
21:38:12 Join lImbus [0] (
21:38:16lImbushi all
21:38:33lImbusamiconn: I forgot to test usb with the external mmc...
21:39:58amiconnI think if rockbox access to both internal and external mmc works, and usb works for internal, then usb will work for external too.
21:40:31lImbusok, I was not sure, but if YOU say... :)
21:41:23amiconnBtw, there is a hardware limitation in the Ondio: If you connect the Ondio to usb with an MMC inserted, you can't access it. You first have to remove it (or connect with no card). Then you can insert and access a card
21:42:14amiconnBoth archos fw and rockbox will tell you :)
21:42:27lImbuswhy ? sound strange. don't think it's fixeable in rockbox, eh ?
21:42:33amiconnNope, sorry
21:42:59amiconnThe thing is, there are 2 protocols to access an MMC, MMC mode and SPI mode.
21:43:57amiconnThe USB bridge uses MMC mode, while rockbox (has to) use SPI mode. You can switch a card from MMC mode to SPI mode with a command, but you cannot switch back wihout power cycling the card
21:44:58amiconnThere would have been a simple solution - a transistor to switch MMC power off & on. Unfortunately that is not part of the archos hw design...
21:45:12lImbusI didn't like mmc/sd-cards from the beginning
21:45:29 Join scott666_ [0] (
21:48:33 Join casualtie [0] (
22:08:14***Saving seen data "./dancer.seen"
22:38:41 Join [IDC]Dragon [0] (
22:38:52amiconnhi again
22:39:06[IDC]Dragonhi there
22:39:25[IDC]Dragonthe Ondio crowd? ;-)
22:40:03lImbusno other owners yet ?
22:40:16[IDC]Dragona handful
22:40:42[IDC]Dragonand the "dark owners"
22:40:52amiconn[IDC]Dragon: I have a question and probably suggestion(s) for solving the ata init problem when returning from the archos charging screen. (This problem also hits on the player)
22:40:55lImbusthat's not much, compared to the size of my hands :-)
22:42:46amiconnThe question first: What would be the problem with always issuing ata_hard_reset()? Increased boot time? Wearing the hd? Possible data loss?
22:43:12[IDC]Dragonboot time, probably
22:44:28[IDC]DragonI remember once we (unnecessarily) resetted bot master and slave, costing significant time
22:44:54amiconnImho there are 2 possible solutions. The coldstart isn't detected by rockbox because the archos firmware initializes all ports at once.
22:45:05[IDC]Dragonthe reset works for both, so it was unnecessary
22:46:36amiconnThe first option involves changing the bootloader to deliberately *deinitialize* the ata port. This would work because the archos fw has to jump to the coldstart address to start rockbox from flash after leaving the charging screen, so the bootloader is executed again
22:46:57 Quit casualtie ("")
22:47:57[IDC]DragonI don't think the bootloader is executed again, why should it?
22:47:58amiconnThe second option would be to do a retry in ata_init() in case the first init sequence doesn't work, issuing ata_hard_reset() before the 2nd try
22:48:18[IDC]Dragonsounds better
22:48:51amiconn(bootloader) How else should the archos firmware restart rockbox *from flash*? It doesn't know anything about the flash structure you established...
22:49:51[IDC]DragonArchos firmware restarts Rockbox? I'm a bit confused
22:50:44[IDC]Dragonah, yes
22:51:16amiconnIf you start the box by inserting the charger while holding F1/-, the archos firmware starts and displays its charging screen. If you then press On, it leaves the charging screen, and if you don't also hold F1/- while doing this, rockbox is started from flash...
22:51:31amiconn...which then panics with ata -31
22:52:16[IDC]Dragonyes, yes, that's the problem
22:52:47[IDC]Dragonso the de-init is not a bad idea
22:53:37[IDC]Dragonvery nice of the Archod firmware, how convenient
22:54:39 Quit methangas (" HydraIRC -> <- Nine out of ten l33t h4x0rz prefer it")
22:55:08amiconnIt's a quite rare situation to get this panic because of the many steps involved, but it is somewhat likely that at least recorder users get this, when they use archos charging for whatever reason
22:56:36amiconnThe bootloader solution probably needs less code, but on the downside requires a full reflash to work
22:57:10[IDC]Dragonyes, that's the downside
22:59:40[IDC]Dragonan ATA reset retry might be a good idea, independently
23:06:38 Join mattzz [0] (
23:08:29amiconn[IDC]Dragon: I wonder why the io_address_detect() return code is also for error. As it is implemented now, it never returns an error, as it simply checks the hw mask. Plus, it could be done earlier, independently of the rest of the init sequence
23:08:38amiconn*is also checked for error
23:12:06[IDC]Dragonah, I once changed that
23:12:47[IDC]Dragonbefore determined by the mask, the address was determined by a rather complex probing
23:13:37[IDC]Dragonwhich probably had failue conditions, too
23:14:08amiconnAh. So this could be moved towards the top, leaving 4 function calls to be put in a separate function, and then doing like this:
23:15:05amiconnif (!coldstart) rc = init_and_check(false);
23:15:33amiconnif (rc != 0) rc = init_and_check(true);
23:15:44amiconnif (rc != 0) return rc;
23:16:13amiconnint init_and_check(bool hard_reset)
23:16:20[IDC]Dragonyes, looks good
23:17:05[IDC]Dragonor: if (!coldstart) rc = init;
23:17:33[IDC]Dragon if (rc != 0) hard_reset(); init();
23:17:55[IDC]Dragonmissing braces above
23:18:07amiconnNot quite. The init function needs the fact whether to do hard reset or not in 2 places.
23:18:15[IDC]Dragonah, ok
23:19:50amiconnI found some more things about the old vs. new player: (1) From the disasm, I got that our new_player check is slightly off.
23:20:43amiconnWe do if (rom_version > 451). Correct is if (rom_version > 449 || rom_version == 116) Yes, strange!
23:21:15[IDC]Dragondifferent subject: for the Ondio, we can remove the ? in the battery symbol
23:21:28[IDC]Dragon(just appeared to me)
23:22:14amiconnProbably, yes. The initial consumption (influencing the battery reading) is not that much higher as on the hd players
23:22:59lImbusgood luck guys, gotta go to bed
23:23:01amiconnThen I tested ata power off. It ran without problems for 20 hours on my box, powering from psu, and currently almost 1 hour from batteries
23:23:13[IDC]Dragonold vs. new player influenced LCD, ATA power pin, any more?
23:23:23 Quit lImbus (" WOW! This IRC Client ownz! HydraIRC -> <-")
23:23:30amiconnNothing more that we know of
23:24:14[IDC]Dragonsaw you mentioning that to Linus, yes. He immediately had problems?
23:24:57amiconnBox switches off as soon as it tries to power the hd.
23:25:17[IDC]Dragonmaybe the power surge
23:25:22amiconnIt cannot do this always, because otherwise it would always immediately power down.
23:25:47amiconnEither there are some bad connections, or maybe it's a hardware revision thing
23:26:26amiconnHis newplayer is one of the early ones, both ROM & archos image version 4.53
23:26:40[IDC]Dragonor recorders, we can switch if the HD should be truely insulated or just sleep (default)
23:27:15[IDC]Dragonwhy not do the same thing for players?
23:27:27amiconnI know. I *can* do the same on my player now
23:27:44[IDC]Dragonthe choice, I mean
23:28:13[IDC]Dragonwith the default being the "classic" way
23:28:32amiconn...which it is anyway, also on recorder
23:29:40amiconnI just would want to dynamically hide the menu option for old players. Although it wouldn't hurt (the power function checks the rom version before doing its work)
23:30:17[IDC]Dragonthat means to un-const it
23:30:41amiconnYes, unfortunately. Would only be needed on the player though
23:31:10[IDC]Dragonlooks even worse
23:31:18[IDC]Dragonwith #if
23:31:28amiconnSome more #ifdef'ing, yes :-/
23:31:57[IDC]Dragonwe introduced a lot already, one more is nothing
23:32:02 Join LinusN [0] (
23:32:14LinusNamiconn: it does *always* power down
23:32:15amiconnhi LinusN
23:32:45amiconnWell, not the very first time. Otherwise you wouldn't be able to use it at all ;-)
23:32:47LinusNas soon as i set the Disk Poweroff setting
23:33:05LinusNit is unable to save the setting on disk
23:33:08amiconnIt powers down as soon as the disk power is cut??
23:33:36LinusNi enable the setting when the disk has spun down
23:33:51LinusNthen it reboots when it tries to spin up
23:34:03LinusNand it can't save the setting
23:34:11[IDC]Dragonnatural protection against an unsuitable setting ;-)
23:34:22amiconnAh. What happens if you set the setting while the disk is not yet spun down?
23:34:25LinusNlucky for me, otherwise it would never be able to boot :-)
23:34:57amiconnAh. Probably better don't try.
23:35:14LinusNyup :-)
23:35:23*[IDC]Dragon proposes the "default setting" boot key combo again
23:35:49LinusN[IDC]Dragon: good idea
23:35:49[IDC]Dragonin german, we call it the grandma key
23:36:00amiconnLinusN: However, you should be able to start again, you merely have to be fast enough to reset the setting before the first spindown
23:36:18amiconnIf you set a large spindown delay, it should be manageable
23:36:31[IDC]Dragonsome TV sets have such a key on the remote, if you're lost in messing everything
23:37:41amiconnLinusN: Any chance to try this on more newplayers, preferably with different rom versions, and ideally one more with your version? I vaguely remember that you once talked about a hardware survey...
23:38:11LinusNwell, kjer has one, and matsl has one
23:38:21LinusNboth work at contactor
23:38:44Bagdermatsl is even joined here atm
23:38:52Bagdermatsl: present?
23:38:54LinusNmatsl: ho ho
23:39:42[IDC]Dragonwhere in the code is that battery status question mark? I can't find it.
23:40:34 Quit gromit` (Read error: 104 (Connection reset by peer))
23:40:49 Join gromit [0] (
23:41:27amiconn[IDC]Dragon: I think it's sufficient to not suppress the first battery reading. The question mark is displayed as long as the battery state is -1
23:41:58[IDC]Dragonyes, I wanted to backtrack how it's handled
23:42:44amiconnThe battery reading or the battery symbol display?
23:43:28[IDC]Dragonfound it, never mind
23:45:37matslLinusN: here!
23:46:17LinusNmatsl: which ROM version is your Player?
23:46:51matslLinusN: huh, don't know.
23:47:19 Quit mattzz ("Client exiting")
23:47:25[IDC]Dragonperhaps your player knows? ;-)
23:48:12matsli'm asking it now.. w8
23:57:04matslversion 5.03
23:57:57amiconnIs that the version rockbox displays under debug, or what archos displays on boot? They may be different (mine are)

Previous day | Next day